3,030 and 1,440 Votes in Songdo, 1,401 and 120 Votes in Gwangju and South Jeolla... The Secret of "Twin Vote Counts" [Reading Science]

"Statistically Impossible?" Suspicions Spread After the Election

Statisticians: "A Natural Coincidence Among Thousands of Combinations"

In the advance voting held in Songdo 1-dong, Incheon, candidate Park Chandae received 3,030 votes, while candidate Yoo Jeongbok received 1,440 votes. Surprisingly, in Songdo 2-dong, the number of votes for both candidates was exactly the same: 3,030 votes for Park and 1,440 votes for Yoo.

A similar situation occurred in Gwangju and South Jeolla Province. In Songjeong 1-dong, Gwangsan-gu, Gwangju, candidate Min Hyungbae received 1,401 votes while candidate Lee Junghyun received 120 votes; the same numbers were recorded in Geumsan-myeon, Goheung-gun, South Jeolla Province. With these cases of so-called "twin polling stations," where the number of votes for candidates matched perfectly, some politicians and social media users raised suspicions, claiming that this is "almost statistically impossible."

On the afternoon of the 3rd, after the voting for the 9th Nationwide Local Elections ended, election clerks were sorting ballots at Jeju Halla Gymnasium. Photo by Yonhap News.

However, statisticians have the opposite interpretation. While the numbers may be eye-catching, mathematically, this is a coincidence that is entirely possible.

"Probability Is Not Zero... Around 0.6% to 0.9%"

Lee Yoon-dong, a professor at Sogang University Business School, explained in an interview with the Korea Science and Technology Media Center (SMCK), "What matters is not comparing the results of just any two regions, but calculating the probability of matching vote counts between areas with similar voting tendencies and number of voters."

For example, in cases like Songdo, where the total number of voters is about 4,470 and the vote ratio between the two candidates is about 2 to 1, the results cluster around a specific segment near the average. According to the central limit theorem, not just any number can appear among thousands of possibilities; similar numbers appear repeatedly.

Professor Lee analyzed that "in this case, the probability that the number of votes for a particular candidate matches exactly in two regions is roughly between 0.6% and 0.9%."

The "Illusion" Created by Thousands of Combinations

A probability of around 1% may seem small. However, in statistics, it is important to consider all possible cases, not just a single event.

Heo Myunghoe, a professor in the Department of Statistics at Korea University, explained that when two people each toss a coin 4,470 times, the probability that the number of heads matches exactly is about 0.9%, according to computer simulation results.

The key point is that the comparison is not limited to just one pair. Assuming there are 137 administrative districts in Incheon, the number of possible pairs is as many as 9,316. Therefore, it is actually natural that some of these thousands of combinations would show identical results.

In Gwangju and South Jeolla Province, this phenomenon can occur even more frequently. This is because a specific candidate’s approval rating is extremely high—around 90%—and there are 393 towns, townships, and neighborhoods, resulting in more than 70,000 possible pairs for comparison. In fact, five pairs of "twin polling stations," where the vote counts matched exactly, were found in Gwangju and South Jeolla Province.

On the 3rd, the 9th nationwide simultaneous local elections day, voters are casting their ballots at the Jamsilbong-dong 4th, 5th, and 6th polling stations set up in the gymnasium of Jamjeon Elementary School in Seoul. Photo by Yonhap News Agency

Numbers Are More Powerful Than Facts

Experts point out that this controversy highlights issues with data interpretation rather than the election results themselves.

Park Hanwoo, a professor at Yeungnam University, said in an interview with the Korea Science and Technology Media Center (SMCK), "It is not a scientific approach to conclude wrongdoing based solely on the rarity of a phenomenon," adding, "Instead of focusing on a single number, we need to examine the overall data structure and context together."

He also stressed, "When numbers labeled as probabilities spread without a clear basis for calculation, they can fuel social distrust. Transparency in data release, expert verification, and the calculation process for probabilities is crucial."

3,030 votes and 1,440 votes, as well as 1,401 votes and 120 votes. At first glance, these may seem like coincidences that are difficult to explain. However, for statisticians, these numbers are not "impossible events," but rather natural probabilistic occurrences that can appear at least once among countless combinations.

This content was produced with the assistance of AI translation services.